Device-free detection of human movement in outdoor environments using Wi-Fi channel state in formation

Abstract

Real-time pedestrian sensing at crossings is essential for Intelligent Transportation Systems (ITS). World Health Organization (WHO) reports that vehicular accidents claim nearly 3,500 lives daily and result in 20 to 50 million injuries annually. Tra- ditional vision-based solutions and wearable devices attempts to solve this challenge however face privacy, cost and deployment challenges. While Wi-Fi channel state in- formation (CSI) has shown promise in indoor environments, its potential for dynamic outdoor environments remains largely unexplored. This paper presents a device-free, CSI-based system for detecting human movement and walking direction in crossings and alert oncoming vehicles to enhance pedestrian safety. A two-stage classification frameworkisadoptedwithAmplitudeandPhase-sensitivefeaturesfilteredwithaadap- tive noise suppression and evaluated using classical and deep learning models. Tested across four outdoor locations and seven subjects, the system demonstrates 95.9% ac- curacy for movement detection and up to 62.4% for directional classification. Results demonstrate the feasibility of deploying low-cost commercial off-the-shelf (COTS) hardware such as ESP32 and Raspberry Pi hardware for scalable, privacy-preserving ITS applications.